agentsclimarketplace

Triage shaping

Skill lherron/agent-metaskills/skills/triage-shaping

Shape and refine a wrkq task by ID. Use when given a task ID to analyze, clarify requirements, and update the task with a structured implementation approach.From its SKILL.md

Install
npx -y skills add lherron/agent-metaskills --skill triage-shaping

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.8 KB, as published. Nobody here has run it

Triage Shaping

Shape an existing wrkq task into a well-defined, actionable specification.

Usage

/triage-shaping <wrkq cat task> OR <task-id> <task-slug>

Process

Step 1: Fetch Task Details if not provided

if the full task was not provided, read the task by ID:

wrkq cat <task-id>

Step 2: Understand the Ask

Before shaping:

  1. Read the task description thoroughly
  2. Explore any referenced files or code
  3. Understand the scope and constraints
  4. Identify what's ambiguous or underspecified

Step 3: Context Grounding (Environment Scan)

Orient to the environment:

  • Scan for existing patterns that inform reasonable defaults
  • Note what's present that the proposal should integrate with
  • Identify constraints the user may not have stated but likely expects

Step 4: Shape the Implementation

Apply prompt-shaping techniques:

  1. Restate in natural language - Articulate what you understand the task wants, making assumptions explicit
  2. Bullet the structure - Break the restatement into scannable components
  3. Propose an iterative grounding loop - Define how to verify the work is complete
  4. Identify decisions needed - Surface any ambiguities requiring user input

Step 5: Iterative Testing

Propose how completion will be verified:

  • Define success criteria - What observable outcome proves the request is fulfilled?
  • Propose a verification method - A concrete check (test, command, assertion)
  • Commit to iteration - Run verification until it passes

Use fully integrated tests, not mocks.

Step 6: Present Shaped Task

Display the refined task to the user:

  • Original task summary
  • Shaped interpretation with explicit assumptions
  • Structured breakdown of work
  • Success criteria and grounding loop
  • Any questions or decisions needed

Step 7: Update Task in wrkq

Once the user approves the shaped approach, offer to:

  1. Update the task description with the shaped specification. Retain the original user text for full historical context:
wrkq edit <task-id>
  1. Mark the task as open (ready for execution):
wrkq set <task-id> --state open
  1. Add a comment documenting the shaping:
wrkq comment <task-id> "Shaped implementation approach - ready for execution"

Example

User: /triage-shaping T-00042

Claude: Let me fetch and analyze this task...

[Runs: wrkq cat T-00042]

**T-00042: Add user preferences API**
- Priority: 2
- State: draft
- Description: "Need some way to save user preferences"

This task is underspecified. Let me shape it:

---

**Shaped Interpretation:**

It sounds like you want to persist user-configurable settings. I'm assuming:
- REST API endpoints (GET/PUT /api/preferences)
- JSON storage per user
- Schema includes: theme, notifications, locale
- Authenticated users only

**Structured Breakdown:**
1. Define preferences schema/types
2. Add database migration for preferences table
3. Implement GET endpoint (fetch current)
4. Implement PUT endpoint (update)
5. Add validation middleware

**Grounding Loop (Iterative Testing):**
I'll write integration tests for both endpoints, iterating until they pass with valid auth and reject without.  I will not use mocks.

**Questions:**
- Should preferences sync across devices?
- Any specific settings beyond the defaults I listed?

---

If this looks correct, I can update T-00042 with this specification and mark it open.

Notes

  • This skill shapes tasks but does NOT execute them
  • The goal is to transform vague requests into actionable specifications
  • Always surface assumptions explicitly
  • The user decides whether to update the task or refine further

Keep looking

Skills are one crate of 627,801. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.